The “What is Entrepreneurship?” section of our website discusses some inherent challenges associated with defining entrepreneurship, and a lack of consensus regarding how to conceptualize it. This elusiveness presents similar challenges for the measurement of entrepreneurship – after all, you cannot measure what you cannot define.
There are no perfect measures of entrepreneurship, but with a meaningful working definition of entrepreneurship and suitable datasets available on which to draw, we can get a fairly comprehensive and reasonably reliable view of the state of entrepreneurship in the American economy.
CAE’s preferred definition of entrepreneurship contains four key elements:
- Entrepreneurs form new companies, with the aim of delivering something new or improved to the market, or by organizing the means of production in a superior way;
- Entrepreneurship refers to the process of starting, developing, and evaluating the viability of a business model, generally around a novelty (product, process, or superior production and delivery);
- The start-up process is temporary in duration – firms either discover a sustainable business model, grow and become successful (including being acquired), or they close their doors; and,
- Start-ups differ from other small or young businesses in that, realized or not, growth is a principle aim for success (as opposed to lifestyle businesses or inherently small business that exist to serve a defined local customer-based).
With these characteristics in mind, economists and other researchers look to three ways for measuring entrepreneurial trends in the economy:
- Study of businesses by firm age – in particular, newly formed businesses or young businesses (often those aged 5 years or less as a proxy for start-ups);
- Study of businesses by growth dynamics – in particular, businesses that achieve high rates of growth; and,
- Study of businesses by innovation – given data limitations, this has generally meant data on young businesses that attract venture capital investments.
One characteristic not included is firm size. Young businesses do start small, but we are principally concerned with young firms that grow – or at least have an aim to.
Finally, it is vital that measurement of entrepreneurship is collected at the level of the entire enterprise. A business establishment is a physical location of economic activity, while a firm is the central controlling entity of a business enterprise. Most firms are single establishment, meaning that they have only one location. But others firms, especially large firms, maintain multiple locations – some even have hundreds or thousands of them.
The distinction between a firm and an establishment is critical – a new Starbucks store or a new Google campus denotes the expansion of an existing firm (establishment birth), while a new locally owned café or a software startup denotes the formation of a new firm (firm birth). However, because data on the former are easier to obtain than the data on the latter, some researchers and other observers conflate the two. We do not recommend this approach.
* * *
The most common method of measuring aggregate entrepreneurship is via the formation of new companies. This is accomplished, primarily, through data sources that track businesses by firm age. Below are four highly reliable, and publicly available, data sources on trends in business formation in the United States.
The premier source of readily available business statistics by firm age in the United States, this series tracks the entire universe of business enterprises that employ at least one person (we typically ignore “firms” without employees, because they are not linked to growth). Data are available annually from the late-1970s through the present across dimensions of firm age, size, geography (U.S., states, and metro areas), and broad industrial sector. The BDS tracks employment, establishment entry and exit, firm entry and exit, and the job creation and destruction associated with the growth and contraction of these firms. The BDS data do lack timeliness, as they are typically published with a two-year lag.
The BLS has recently developed an experimental series, which is similar to the BDS, though it has a wider universe of businesses. The series tracks firms and establishments annually, and the data are available beginning in 1994 through the present (there is typically a one-year lag time). The BED data are available by firm age, size, geography (U.S. and states), and industry (by two-digit NAICS code).
An annual supplement to the quinquennial (every five years) Survey of Business Owners (SBO), this is a statistically representative sample of employer businesses in the United States. Beginning in 2014 and continuing through the present (with a two-year lag), this series reports businesses counts, employment, revenues, and wages by 40 characteristics of businesses (e.g. aspirations, financing, intellectual property) and 11 characteristics of business owners. To capture newer businesses (“entrepreneurs”), measures are disaggregated by firm age. The data are available by geography (U.S., states, and the 50 largest metros) and industry (by two-digit NAICS code).
Tabulated annually, the KISA relies on a three-pronged approach to generate a broad-based measure of new business formation in the United States. The first relies on the U.S. Census Bureau’s Current Population Survey, a statistically representative sample of U.S. households. Because respondents participate in the survey over time, KISA is able to identify individuals who are business owners in one year but were not in the year prior. A subset of this data is extracted to distinguish “opportunity” entrepreneurs (those starting businesses who were previously in paid employment or in school) from “necessity” entrepreneurs (those that started businesses from unemployment). The final component of the three-part index is the headline figure for business formations from the BDS described above.
* * *
One major limitation of the firm formation approach is that it doesn’t distinguish between businesses that intend to grow from those that are either lifestyle businesses or intend to stay small, serving a local customer base. Several sources provide metrics on those.
This is a firm-level database of the entire business sector in the United States, which allows authorized researchers to calculate the distribution of firm growth rates by a range of industry, geography, age, or size dimensions. Simply put, this is by far the best available dataset to study business demography and growth dynamics in the United States. The main constraint, however, is that access is highly restricted (for academic researchers only) and the publication of results is restricted, so most researchers opt for reduced, publicly-available versions such as the BDS (the LBD firm-level dataset is used to produce various aggregates that comprise the BDS).
Inc. Magazine produces an annual list of the fastest growing private companies in America. To qualify, companies must meet the following criteria: be privately and independently owned, based in the United States, and have had revenues of no less than $2 million the year prior and no less than $100 thousand three years prior. Eligible companies with the highest verified revenue growth over this three-year period make the list, and are ranked accordingly. There are some limitations to this approach, including the fact that some fast-growing companies have been left out. Even so, these lists present valuable information on some of the fastest growing businesses in America each year.
A recently inaugurated annual index, the KIGE takes a three-pronged approach to estimate the activity of “growth entrepreneurship” in the American economy. The first component comes from the BDS dataset described above, and tabulates the average growth of a start-up cohort over its first five years. The second factor, again using the BDS dataset, is the rate of “scale-up” firms in the economy – the share of firms aged under ten years that started small and grew to at least 50 employees. The third component utilizes the Inc. 500 List described above and calculates the “density” of these firms relative to total firms, after normalizing for growth thresholds. Once the individual metrics are compiled, an index is calculated for the United States, the 50 states, and 40 metropolitan areas, from 2008 to present (with about a one-year lag).
* * *
Finally, many researchers utilize private venture capital databases to assess funding trends – and, therefore, company trends – of some of the country’s highest potential businesses. The argument in favor of using venture capital investments as a proxy for early-stage innovative activity – or also as a proxy for the prevalence of very high-growth potential companies – is that companies receiving these investments have been deemed by professional investors to have the potential to grow exponentially within a short amount of time (less than 10 years).
The argument against using these databases rests on the fact that venture capital investments don’t necessarily translate to actual growth. Additionally, only a very small percentage (fewer than 2 percent) of all new companies ever receive venture capital funding.
Even so, these are useful proxies for activity of very high potential companies, and a good way to determine which emerging technologies professional investors are backing.
Some prominent venture capital databases, which require paid subscriptions with private data vendors, include: Pitchbook, CB Insights, VentureSource, Thomson Reuters, Preqin, Capital IQ, Crunchbase, and Mattermark.